Burnett Measurements for the Difluoromethane + Carbon Dioxide

The apparatus was calibrated using helium, and its performance was confirmed by measurements for pure R32. The second and third virial coefficients we...
0 downloads 0 Views 78KB Size
876

J. Chem. Eng. Data 2002, 47, 876-881

Burnett Measurements for the Difluoromethane + Carbon Dioxide System G. Di Nicola,*,† F. Polonara,† and R. Stryjek‡ Department of Energetics, University of Ancona, Ancona, Italy, and Institute of Physical Chemistry, Polish Academy of Sciences, Warsaw, Poland

As part of our ongoing research program on PVTx measurements using the isochoric and Burnett methods for hydrofluorocarbon + hydrocarbon and/or inorganic compound systems, this work reports on the experimental results for the difluoromethane (R32) + carbon dioxide (CO2) system obtained by the Burnett method. The apparatus was calibrated using helium, and its performance was confirmed by measurements for pure R32. The second and third virial coefficients were derived, and a good consistency was found after comparison with data in the literature. A new experimental procedure for mixture measurements was developed, on the basis of the gas chromatographic analysis of the sample collected during expansions. The gas chromatograph was calibrated precisely using samples prepared by the gravimetric method. PVTx measurements were performed for the binary R32 + CO2 system at T ) (303, 313, 323, 333, and 343) K. The second and third cross virial coefficients were derived from the experimental results. No experimental data could be found in accessible literature sources on the PVTx for the R32 + CO2 system. Results were compared with the prediction obtained using REFPROP 6.01.

Introduction Carbon dioxide has received much attention for industrial applications. It is also an attractive compound for research because of its convenient critical temperature and thermal stability, which explains the fairly abundant and accurate literature providing thermophysical data for CO2.1 It is neither flammable nor toxic, is inexpensive and is widely available. One of the main disadvantages of this natural fluid is that the plant components need to be designed for much higher pressures; typically 150 bar is the maximum operating pressure. Difluoromethane is a hydrofluorocarbon used in the refrigeration industry. In particular, it is an important alternative refrigerant as a constituent in several binary or ternary mixtures intended as a substitute for difluorochloromethane (R22), and its thermophysical properties have been studied extensively in the past decade. The main limit of this refrigerant fluid lies in its flammability. This alternative refrigerant has a zero ODP but a far from negligible GWP, so its combination with a low-GWP and nonflammable compound such as CO2 is particularly promising. Nevertheless, publications on the PVTx properties of systems containing carbon dioxide as one of their components and a hydrofluorocarbon as the other are rare. Studying the combination of carbon dioxide and difluoromethane thus provides important information for future applications. Moreover, no experimental results have been published to our knowledge on the PVTx properties of this specific binary system. This work presents and discusses the PVTx resulting from Burnett measurements. Experimental Section Reagents. CO2 and R32 were supplied by Sol SpA and Ausimont SpA, respectively; the purity was checked by gas * To whom correspondence should be addressed. † University of Ancona. ‡ Polish Academy of Sciences. Visiting scientist at CNR-ITEF, Padua, Italy

Figure 1. Schematic view of the experimental apparatus: 1, nitrogen reservoir; 2, vacuum pump (Vacuubrand, mod. RZ2); 3, precision pressure controller (Ruska, mod. 3981); 4, gas lubricated dead weight gauge (Ruska, mod. 2465); 5, vibr. cylinder pressure gauge (Ruska, mod. 6220); 6, digital temperature indicator (Corradi, RP 7000); 7, electronic null indicator (Ruska, mod. 2416); 8, stirrer; 9, heater; 10, cooling coil connected with an auxiliary bath; 11, differential pressure transducer (Ruska, mod. 2413); 12, measurement chamber (VA); 13, expansion chamber (VB); 14, magnetic recirculating pump; 15, Pt resistance thermometer (Tersid, Pt 100); 16, vacuum pump for VB (Vacuubrand, mod. RZ2); 17, charging fluid reservoir; 18, Pt resistance thermometer (Hart Scientific, Pt 25); 19, digital pressure indicator (Ruska, mod. 7000); V1, V2, V3, V4, constant volume valves.

chromatographic analysis, using a thermal conductivity detector. It was found to be 99.99% for the CO2 and 99.98% for the R32 on an area-response basis. Apparatus. A diagram of the apparatus is shown in Figure 1. It is the same as the one described elsewhere2-4 and used with only minimal modifications. It consists of

10.1021/je015537m CCC: $22.00 © 2002 American Chemical Society Published on Web 05/22/2002

Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002 877 Table 1. Experimental Pressures Measured during Burnett Expansions of R32 P/kPa series 1 series 2 series 3 series 4 series 5 series 6a series 7a series 8 series 9 T/K ) 303.15 T/K ) 303.16 T/K ) 313.15 T/K ) 313.16 T/K ) 323.16 T/K ) 323.16 T/K ) 333.16 T/K ) 343.16 T/K ) 343.16 1548.2 1119.8 788.6 545.5 373.0 253.3 171.1 115.2

a

1783.7 1313.8 935.6 652.0 448.1 305.2 206.7 139.4

1730.8 1249.4 878.4 607.3 415.1 281.6 190.1 128.0

2363.1 1786.3 1293.4 911.2 630.8 431.5 292.9 197.9 133.3

2291.1 1685.3 1200.0 835.9 574.4 390.9 264.2 177.9

3110.5 2444.3 1816.3 1301.9 911.3 628.0 428.4 290.1 195.8 131.6

2123.3 1526.7 1071.3 739.7 504.9 342.1 230.6 155.1

4198.0 3335.6 2503.2 1807.8 1273.4 881.3 603.4 409.5 276.7 186.4 125.2

4365.7 3520.5 2667.7 1940.4 1372.7 953.2 653.3 444.2 300.5 202.5 136.2

Data obtained with N ) 1.495 70.

Table 2. Second and Third Virial Coefficients for R32a series

T/K

B/cm3‚mol-1

C/cm6‚mol-2

F(1)a/mol‚dm-3

abs(dP)/kPa

100(P - Pcalc)P-1/%

1 2 3 4 5 6b 7b 8 9

303.15 303.16 313.15 313.16 323.16 323.16 333.16 343.16 343.16

-284.3 -283.5 -257.5 -257.4 -238.7 -238.3 -220.8 -203.2 -202.3

24 420 23 940 20 650 19 890 19 570 18 980 17 380 16 070 15 760

0.772 70 0.936 77 1.294 02 0.831 16 1.852 09 1.129 25 0.951 37 2.465 54 2.680 47

0.1 0.0 0.1 0.1 0.2 0.2 0.2 0.4 0.6

-0.01 0.01 -0.01 -0.02 -0.06 -0.04 -0.05 0.04 0.06

a

F(1) denotes regressed initial density. b Data obtained with N ) 1.495 70.

Table 3. Second and Third Virial Coefficients for CO2a series

T/K

B/cm3‚mol-1

C/cm6‚mol-2

F(1)/mol‚dm-3

abs(dP)/kPa

100(P - Pcalc)P-1/%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

283.14 283.14 283.14 303.15 303.15 303.15 303.15 313.15 323.16 323.16 333.63 333.63 343.15 343.15 363.15 363.15 363.15

-136.2 -136.6 -136.4 -118.9 -118.1 -119.2 -118.9 -109.8 -103.6 -103.7 -95.0 -94.5 -88.7 -90.2 -79.6 -78.4 -75.9

4390 4580 4890 4870 4690 4980 4860 4600 4660 4680 4040 4000 3920 4250 4130 3670 3210

2.246 27 2.534 39 2.570 16 3.280 11 3.118 60 3.274 10 3.301 30 2.808 84 2.812 53 2.878 13 2.495 94 2.541 26 2.344 21 2.475 49 2.164 73 2.099 60 2.147 79

0.3 0.2 0.1 0.1 0.2 0.0 0.0 0.1 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.2 0.4

-0.30 -0.07 -0.03 -0.09 -0.04 -0.02 -0.02 -0.01 -0.03 -0.09 -0.09 -0.25 0.02 0.03 0.00 0.01 -0.03

a

F(1) denotes regressed initial density.

two pressure vessels, the measurement chamber, VA, and the expansion chamber, VB, with a volume of approximately (70 and 35) cm3, respectively, and some auxiliary systems for filling and mixing the compounds in the Burnett vessels and for controlling and measuring the pressure and temperature. The spherical shape of the vessels simplifies the evaluation of the distortion coefficients as a function of temperature and pressure. Moreover, the ratio of outer and inner diameters of both spheres is the same. The relative volume change with pressure is consequently identical for the two vessels. The vessels are made of Invar, because of its excellent corrosion resistance and low thermal expansion coefficient. The constant volume valves have been redesigned to ensure thermal compensation of the packing and valve body. The four-valve arrangement enables the vessels VA and VB to be filled or emptied separately and, in addition to the expansion experiment, allows for the compounds in the Burnett vessels to be mixed using a magnetic recirculating pump.

The packing surfaces of valves V1 and V4 are exposed to the expansion volume VB, and those of valves V2 and V3 are exposed in the opposite direction of the volumes VA and VB. Thus, the principal volume VA and its fixtures are allmetal, to prevent contact between the sample gas in vessel VA and the Teflon packing throughout the lengthy Burnett experiment, except for the time it takes to reach thermal equilibrium after expansion. The measurement vessel is connected to a diaphragmtype differential pressure transducer (Ruska Model 2413), which is coupled to an electronic null indicator (Ruska Model 2416). The pressure is regulated by a precision pressure controller (Ruska Model 3981), while a digital pressure indicator (Ruska Model 7000) is used to measure pressures. Nitrogen is used to balance the sample gas pressure, and the nitrogen circuit consists of a reservoir, expansion vessels, and pressure regulating systems. The vessels are immersed in a thermostatic bath filled with about 45 L of an ethylene glycol and water mixture.

878

Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002 Table 4. Experimental Pressures Measured during Burnett Expansions and Regressed Compressibility Factors of the R32 (1) + CO2 (2) System T/K ) 303.15 Series 1a x1 ) 0.1667

Series 2 x1 ) 0.3192

Series 3 x1 ) 0.4576

Series 4 x1 ) 0.6462

P/kPa

z

P/kPa

z

P/kPa

z

P/kPa

z

4639.0 3532.1 2568.4 1813.5 1256.8 860.2 583.9 394.0 265.0

0.6841 0.7797 0.8487 0.8970 0.9305 0.9534 0.9687 0.9784 0.9851

4077.7 3120.0 2275.8 1611.5 1118.7 766.9 520.4 351.8 237.2 159.4

0.6786 0.7762 0.8463 0.8958 0.9295 0.9526 0.9663 0.9766 0.9841 0.9888

3004.2 2220.2 1584.8 1105.5 760.2 518.3 351.1 236.7 159.1

0.7486 0.8270 0.8824 0.9202 0.9459 0.9639 0.9762 0.9836 0.9883

2353.4 1722.4 1221.6 849.0 582.7 396.5 267.4 180.2

0.7697 0.8420 0.8927 0.9274 0.9515 0.9678 0.9758 0.9832

T/K ) 313.15

Figure 2. Deviations between the second virial coefficients calculated according to the Tsonopoulos method12 and the experimental second virial coefficients for R32 against temperature: b, present work; O, ref 7; ], ref 8; 0, ref 9; 3, ref 10; 4, ref 11.

Series 5 x1 ) 0.3220

Series 6 x1 ) 0.4628

Series 7 x1 ) 0.6312

Series 8 x1 ) 0.8029

P/kPa

z

P/kPa

z

P/kPa

z

P/kPa

z

4222.1 3163.9 2280.5 1602.3 1107.7 757.4 513.9 347.1 233.7 157.0

0.7154 0.8013 0.8634 0.9068 0.9371 0.9577 0.9714 0.9808 0.9872 0.9914

3251.3 2387.2 1696.6 1181.5 811.0 552.2 373.6 251.9 169.3

0.7618 0.8361 0.8883 0.9247 0.9488 0.9657 0.9767 0.9843 0.9890

2367.6 1705.9 1197.7 827.5 565.2 383.4 259.0 174.1

0.8042 0.8662 0.9091 0.9390 0.9587 0.9720 0.9816 0.9866

2294.3 1676.8 1187.9 825.8 566.8 385.7 261.0 175.9

0.7708 0.8421 0.8918 0.9268 0.9508 0.9673 0.9784 0.9859

T/K ) 323.15 Series 9 x1 ) 0.3108

Figure 3. Deviations between the third virial coefficients calculated according to eq 3 and the experimental third virial coefficients for R32 against temperature: 9, present work; ], ref 8; 3, ref 10.

The temperature of the bath is kept constant by means of a system with a PID device, piloted by a computer to which the temperature measurement system is also connected. The temperature control and acquisition system relies on two platinum resistance thermometers calibrated according to ITS 90 at the Istituto Metrologico G. Colonnetti (IMGC), of Turin. In particular, for temperature measurements, a Hart Scientific Pt 25 Ω resistance thermometer (mod. Hart 5680) is used, while for control purposes a Tersid Pt 100 Ω resistance thermometer is used. Both the thermometers are connected to a digital temperature indicator (Corradi, RP 7000). The Burnett constant, N, defined as the ratio of the volume of cell A and the sum of the volumes of cells A and B at zero pressure, was determined by means of gaseous helium measurements (since its thermodynamic properties are well established). After taking measurements at several isotherms, the constant was found to be N ) 1.494 85 ( 0.0001; after servicing the setup, we found the Burnett constant to be N ) 1.495 70 ( 0.0001 and N ) 1.496 91( 0.0001. These results are marked with a footnote in Tables 1, 2, and 4. The Burnett constant found reproduces the pressures of helium with absolute average deviations, AAD ) 0.418 kPa or AAD ) 0.142, in percent.

Series 10 x1 ) 0.3487

Series 11 x1 ) 0.5709

Series 12 x1 ) 0.5866

P/kPa

z

P/kPa

z

P/kPa

z

P/kPa

z

5124.0 3880.7 2818.6 1991.5 1381.6 947.0 643.9 435.5 293.5 197.3 132.3

0.6908 0.7820 0.8491 0.8968 0.9299 0.9529 0.9685 0.9791 0.9864 0.9915 0.9934

3833.7 2791.2 1975.1 1371.6 940.7 639.9 432.9 291.8 196.1 131.6

0.7764 0.8450 0.8938 0.9278 0.9513 0.9673 0.9781 0.9855 0.9904 0.9935

3270.9 2394.7 1700.5 1183.7 813.1 553.6 374.8 252.7 169.9

0.7641 0.8363 0.8877 0.9237 0.9484 0.9654 0.9769 0.9846 0.9898

3409.1 2514.8 1794.7 1253.3 862.8 588.3 398.7 269.0 180.9 121.5

0.7473 0.8241 0.8791 0.9177 0.9444 0.9626 0.9751 0.9834 0.9888 0.9928

T/K ) 333.15 Series 13 x1 ) 0.3222

Series 14 x1 ) 0.4489

Series 15 x1 ) 0.5791

Series 16 x1 ) 0.7372

P/kPa

z

P/kPa

z

P/kPa

z

P/kPa

z

2456.6 1711.9 1177.0 801.8 543.0 366.0 246.2 165.3

0.8842 0.9210 0.9466 0.9640 0.9759 0.9834 0.9889 0.9925

4388.0 3235.9 2310.2 1614.2 1111.7 758.4 514.1 346.9 233.5 156.8

0.7457 0.8220 0.8772 0.9162 0.9433 0.9619 0.9747 0.9833 0.9891 0.9930

3522.3 2566.3 1817.2 1262.6 866.3 589.5 398.8 268.9 180.8 121.4

0.7739 0.8429 0.8922 0.9266 0.9504 0.9667 0.9777 0.9854 0.9906 0.9942

2877.4 2082.6 1468.4 1017.3 696.8 473.6 320.2 215.7 145.0

0.7889 0.8535 0.8996 0.9317 0.9539 0.9691 0.9795 0.9864 0.9912

T/K ) 343.15 Series 17 x1 ) 0.3210

Series 18 x1 ) 0.4956

Series 19 x1 ) 0.6856

Series 20 x1 ) 0.7132

P/kPa

z

P/kPa

z

P/kPa

z

P/kPa

z

4093.8 2922.3 2041.3 1405.5 958.4 649.6 438.1 295.0 198.2 132.9

0.8229 0.8781 0.9169 0.9437 0.9619 0.9746 0.9827 0.9889 0.9931 0.9959

3340.7 2377.3 1657.7 1140.2 777.0 526.4 355.1 238.9 160.4 107.6

0.8293 0.8822 0.9195 0.9454 0.9631 0.9753 0.9835 0.9892 0.9930 0.9952

2253.8 1580.8 1091.6 745.9 506.2 341.9 230.2 154.7

0.8664 0.9084 0.9377 0.9579 0.9716 0.9810 0.9873 0.9917

3160.6 2276.6 1600.4 1106.7 757.0 514.1 347.5 234.1 157.3

0.7994 0.8607 0.9045 0.9350 0.9560 0.9705 0.9805 0.9874 0.9922

a

Data obtained with N ) 1.496 91.

Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002 879 Table 5. Experimental and Calculated Second and Third Virial Coefficients for the R32 (1) + CO2 (2) System Bmexp

Bmcalc

dBm

Cmexp

Cmcalc

dCm

cm3‚mol-1

cm3‚mol-1

cm6‚mol-2

cm6‚mol-2

cm6‚mol-2

-133.3 -151.1 -171.1 -204.1 -141.6 -160.1 -186.2 -216.9 -131.2 -135.5 -164.2 -166.5 -123.2 -137.5 -154.0 -176.6 -110.8 -128.4 -152.3 -168.0

0.3 -0.3 -0.2 0.0 -0.5 2.3 1.6 -2.2 0.9 0.4 -0.8 -0.8 1.2 0.9 0.2 -1.9 0.3 -4.2 -5.3 6.5

5790 6990 8430 11680 7280 7620 10060 16350 6620 7280 10640 10870 6520 7940 9900 13790 5130 8250 12160 12020

5850 6940 8420 11710 6370 8210 11190 15010 6710 7180 10620 10900 6170 8110 10420 13320 5500 8010 11260 13070

-60 50 10 -30 910 -590 -1130 1340 -90 100 20 -30 350 -170 -520 470 -370 240 900 -1050

T/K

x1

cm3‚mol-1

303.15

0.1667 0.3192 0.4576 0.6462 0.3220 0.4628 0.6312 0.8029 0.3108 0.3487 0.5709 0.5866 0.3222 0.4489 0.5791 0.7362 0.3210 0.4956 0.6856 0.7927

-133.0 -151.5 -171.3 -204.0 -142.1 -157.8 -184.5 -219.1 -130.3 -135.0 -165.0 -167.3 -122.0 -136.6 -153.9 -178.5 -110.5 -132.5 -157.6 -161.4

313.15

323.15

333.15

343.15

Table 6. Smoothed Second and Third Virial Coefficents for the R32 (1) + CO2 (2) System B11

B12

B22

C111

C112

C122

C222

T/K

cm3‚mol-1

cm3‚mol-1

cm3‚mol-1

cm6‚mol-2

cm6‚mol-2

cm6‚mol-2

cm6‚mol-2

303.15 313.15 323.15 333.15 343.15

-283.9 -257.5 -238.5 -220.8 -202.8

-154.5 -147.4 -137.6 -129.6 -111.7

-118.8 -109.8 -103.7 -95.1 -89.4

24180 20270 19280 17380 15920

7415 10640 12220 13510 12420

6930 4550 4890 3340 2460

4860 4600 4670 4020 4090

Experimental Procedure, Gas Chromatographic Calibration and Experimental Uncertainties Experimental Procedure. To measure the system components, the classical Burnett experimental procedure was followed. Initially, the first vessel was filled with the sample and its temperature and pressure were measured. Then, after evacuating the second vessel, the expansion valve was opened. Once the pressures between the vessels had equalized, the second vessel was isolated and evacuated again. This procedure was repeated until low pressures were achieved. In the mixture-charging procedure, the two vessels were separately filled with different compounds. Then the fluids were mixed together with the aid of the magnetic pump while the expansion valve was kept open. During the first expansion, a sample of the mixture was collected and used for gas chromatographic analysis of the mixture’s composition. Gas Chromatographic Calibration. The mixtures for the calibration were prepared in stainless steel bottles of about 250 cm3 capacity; the mass of charged components, depending on the composition, was varied from 8 to 15 g and was weighed with an analytical balance to (0.1 mg. On that basis, the molar fraction of the sample was found. Next, the prepared samples were connected on-line to the gas chromatograph, and after the sample was preheated to about 323 K (for the sample homogenization), a series of gas chromatographic analyses was performed. The reproducibility of the gas chromatograph for each sample was within 0.1%. The results of the analysis for the series of samples covering a wide range of composition were next regressed to find the relation between ratio of area and composition in molar fraction. A third-degree polynomial expression, obtained by forcing it also to the points corresponding to pure compounds, gives us the unknown composition when its peak area ratio is measured. The

composition of the mixtures was found from the gas chromatographic analysis performed for the same gas chromatograph parameters and applying the expression found. For the analysis, the samples collected from the first Burnett expansions were used. Further details of the method are given in ref 5. Experimental Uncertainties. The uncertainty in the temperature measurements is due to the thermometer and any instability of the bath. The stability of the bath was found to be better than (0.015 K, and the uncertainty of the thermometer was found to be better than (0.010 K in our temperature range. The total uncertainty in the temperature measurements was thus lower than (0.03 K. The uncertainty in the pressure measurements is due to the transducer and null indicator system, and to the pressure gauges. The digital pressure indicator (Ruska, mod. 7000) has an uncertainty of (0.003% of the full scale. The total uncertainty in the pressure measurement is also influenced by temperature fluctuations due to bath instability and was found to be lower than (1 kPa. The uncertainty of the mixture’s composition was found to be always lower than 0.5% in mole fraction, based on the calibration with samples prepared gravimetrically and on the reproducibility of the gas chromatograph (Carlo Erba Mega Series 5380 with a thermal conductivity detector). Results for R32 and CO2 For R32, 81 experimental points along 5 isotherms (9 sets in all) were collected in the temperature range from 303 K to 343 K and for pressures from 125 kPa up to 4370 kPa. For CO2, 234 experimental points along 7 isotherms (17 sets altogether) were collected in the temperature range from 283 K to 363 K and in the pressure range from 11 kPa to 5700 kPa. The experimental data for R32 are shown in Table 1 while the experimental data for CO2 are reported elsewhere.6

880

Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002

Figure 4. Second virial coefficients for the R32 (1) + CO2 (2) system against the mole fraction at the following temperatures, T: O, 303.15; 0, 313.15; 4, 323.15; 3, 333.15; ], 343.15 K. The lines represent the correlated values.

Figure 5. Third virial coefficients for the R32 (1) + CO2 (2) system against the mole fraction at the following temperatures, T: O, 303.15; 0, 313.15; 4, 323.15; 3, 333.15; ], 343.15 K.

The experimental PVT measurements were used to derive the second, B, and third, C, virial coefficients of the virial equation

P)

(

)

C RT B 1+ + 2 V V V

(1)

In the regression, each run was treated separately and (dP)2 was used as an objective function applying the Burnett constant from the helium calibration. The pressure distortion of the Burnett cells was considered, as explained elsewhere.2-4 Defining the average absolute deviation in pressure as N

AAD )

∑abs(dP)/N

(2)

i)1

the AAD ) 0.239 kPa for R32 and AAD ) 0.163 kPa for CO26 were found, well within the estimated experimental uncertainty. The second and third virial coefficients for R32 and for CO2 are shown in Tables 2 and 3, together with the pressure deviations from the fit that proved to be randomly distributed. A systematic deviation in pressure is not evident for both compounds, enabling us to say that the adsorption does not affect the results. In Figures 2 and 3, the second and third virial coefficients for R32 are plotted and compared with values in the literature.7-11 In Figure 2, the derived second virial coefficients were compared with the values calculated by means of the Tsonopoulos correlating method,12 while, in Figure 3, the third virial coefficients were compared with the ones calculated with the empirical expression

C ) A1 + A2(Tr - 1) + A3(Tr - 1)2

(3)

where A1 ) 14 565.3408; A2 ) -49 881.8850; and A3 ) 105 331.4976. For R32, there is a good consistency with most of the literature data for the second virial coefficient (refs 7, 8, 10, and 11), while the data in ref 9 are systematically lower. For the third virial coefficients, the results (obtained in a different temperature region) clearly follow a common temperature trend with data in the literature,8,10 even if quite big deviations from the trend are evident while comparing data from various sources. For comparison, we

Figure 6. Deviations in pressure produced by the REFPROP 6.0115 at the following temperatures, T: O, 303.15; 0, 313.15; 4, 323.15; 3, 333.15; ], 343.15 K.

only considered literature sources in which original data were reported. For CO2, all the comparisons with the literature are reported elsewhere;6 hence, we only report a brief summary of the results here. For the second virial coefficient, the present experimental results are consistent with the literature sources.1 Also, the Tsonopoulos correlating method13 represents the experimental data well. For the third virial coefficients, the experimental data in the literature are randomly scattered within 25% around values resulting from the Orbey and Vera correlation;14 our results lay well within that limit. Results for the Mixtures For the R32 + CO2 system, 185 experimental points along 20 sets and 5 isotherms were collected within the temperature range from 303 to 343 K and the pressure range from 110 kPa to 5100 kPa. The experimental findings, together with the compressibility factor values, are given in Table 4. The virial coefficients for the mixtures were found by applying the same procedure as that for the pure compounds. The values of the second and third virial coefficients (Table 5), along with the virial coefficients for the pure compounds (smoothed as a function of reduced temperature), were used to derive cross virial coefficients. The results are shown in Table 6. The second cross virial coefficients were calculated from the expression

Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002 881 n

Bm )

n

∑∑B x x

ij i j

(4)

i)1 j)1

for each experimental datum point. Next, the B12 values found for each temperature were averaged; the averaged B12 values, given in Table 6, were used to calculate the deviations of Bm from the experimental values shown in Table 5. The third cross virial coefficients were calculated from the defining equation n

Cm )

n

performance of the apparatus was checked by means of measurements for R32 and CO2, and a good consistency was observed between the virial coefficients obtained and those reported in the literature. The R32 + CO2 system was studied over five isotherms, and second and third cross virial coefficients were derived. Over the entire temperature range, the second virial coefficients showed positive (while the third virial coefficients showed negative) deviations from ideality. Our results proved consistent with the prediction obtained using REFPROP 6.01.

n

∑∑ ∑ C

ijkxixjxk

(5)

i)1 j)1 k)1

Bearing in mind that two-third cross virial coefficients (namely C112 and C122) represent third virial coefficients for each mixture’s composition, their values could not be calculated directly for each datum point, so they were calculated as average values by combining data for each temperature. The averaged values of the third cross virial coefficients are included in Table 6. Using these values, the third virial coefficients for the mixtures were recalculated for each point. The results, together with the deviations, are also included in Table 5. The greatest random error in Bm was found at T ) 343.15 K with AAD ) 4.08 cm3‚mol-1, while at T ) (303.15, 313.15, 323.15, and 333.15) K the average absolute deviation was found to be AAD ) 0.91 cm3‚mol-1. This is presumably due to a possible slightly poorer thermal stability at higher temperatures while a (water + ethylene glycol) mixture is used as the bath fluid. The overall AAD for Bm was evaluated at 2.5 cm3‚mol-1, while the AAD for Cm was 399 cm6‚mol-2, corresponding to an average deviation of around 10%. These values are acceptable if compared with numerous experimental data, obtained in reduced temperature ranges below or close to 1, reported in the compilation.14 The second and third virial coefficients for the system being plotted against the mole fraction in Figures 4 and 5, respectively, show positive and negative deviations from ideality, defined here as

B12 ) (B11 + B22)/2

(6)

C112 ) (2C111 + C222)/3

(7)

C122 ) (C111 + 2C222)/3

(8)

Because of the total lack of data in the literature on the superheated vapor region for this system, the experimental findings were compared with the prediction obtained with the REFPROP 6.0115 software. The calculations were performed at the highest experimental pressures of each run, adopting regressed values for F(1), Bm, and Cm. The differences in pressure are given in Figure 6, showing a consistency within 1%; the figures were systematically below those resulting from our experiment. Conclusions This work reports on experimental results for the R32 + CO2 system obtained with the Burnett method. The

Acknowledgment The authors are grateful to Filippo Giorgi for his kind help. Literature Cited (1) Angus, S.; Armstrong, B.; de Reuck, K. M. International Thermodynamic Tables of the Fluid State, vol. 3, Carbon Dioxide; Pergamon Press: Oxford, 1976. (2) Glowka, S. A Burnett Apparatus for Accurate Determination of the Volumetric Properties of Gases. Naucznaja apparatura 1988, 3, 79-84. (3) Glowka, S. Determination of Volumetric Properties of Ammonia Between 298 and 473 K and Carbon Dioxide Between 304 and 423 K Using the Burnett Method. Polym. J. Chem. 1990, 64, 699709. (4) Glowka, S. Volumetric Properties of Ammonia + Argon, + Helium, + Methane, and + Nitrogen Mixtures Between 298 and 423 by the Burnett- Isochoric Method. Fluid Phase Equilib. 1992, 78, 285-296. (5) Bobbo, S.; Stryjek, R.; Elvassore, N.; Bertucco, A. A Recirculation Apparatus for Vapor-Liquid Equilibrium Measurements of Refrigerants Binary Mixtures of R600a, R134a and R236fa. Fluid Phase Equilib. 1998, 150-151, 343-352. (6) Di Nicola, G.; Giuliani, G.; Polonara, F.; Stryjek, R. PVTx Measurements for the R125+CO2 System by the Burnett Method. Fluid Phase Equilib. 2002, 199, 163-176. (7) Bignell, C. M.; Dunlop, P. J. Second Virial Coefficients for Fluoromethanes and their Binary Mixtures with Helium and Argon. J. Chem. Eng. Data 1993, 38, 139-140. (8) Defibaugh, D. R.; Morrison, G.; Weber, L. A. Thermodynamic Properties of Difluoromethane. J. Chem. Eng. Data 1994, 39, 333-340. (9) Quian, Z. Y.; Nishimura, A.; Sato, H.; Watanabe, K. Compressibility Factors and Virial Coefficients of Difluoromethane (HFC32) Determined by Burnett Method. JSME Int. J. 1993, 36, 665670. (10) Sato, T.; Sato, H.; Watanabe, K. PVT Property Measurements for Difluoromethane (HFC-32). J. Chem. Eng. Data 1994, 39, 851854. (11) Zhang, H. L.; Sato, H.; Watanabe, K. Second Virial Coefficients for R-32, R-125, R-134a, R-143a, R-152a and their Binary Mixtures. Proceedings of the 19th International Congress of Refrigeration, The Hague, The Netherlands, 1995; Vol. IVa, pp 622-629. (12) Tsonopoulos, C. An empirical correlation of second virial coefficients. AICHE J. 1974, 20, 263-272. (13) Orbey, H.; Vera, J. H. Correlation for the third virial coefficient using Tc, Pc and ω as parameters. AICHE J. 1983, 29, 107-113. (14) Dymond, J. H.; Smith, E. B. The Virial Coefficients of Gases; Clarendon Press: Oxford, 1969. (15) McLinden, M. O.; Klein, S. A.; Lemmon, E. W.; Peskin, A. P. NIST Thermodynamic and transport properties of refrigerants and refrigerant mixtures (REFPROP), version 6.01; Boulder, CO, 1998.

Received for review November 19, 2001. Accepted April 21, 2002. This work was supported by the Italian Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica.

JE015537M